Obesity is a significant public health problem. Preventing obesity is much more effective than treating it and once obesity is established it is very likely to persist throughout the lifespan. Prevention programs have had limited success and efforts have therefore increasingly focused on younger and younger age groups. By age 2 years, 27 percent of children are already overweight or obese and the rate of weight gain in the first year of life is an independent and robust risk factor for obesity later in childhood and into adulthood, independent of prenatal factors or birth weight. Efforts have therefore begun to focus on the rate of weight gain in infancy. The mechanism of rapid weight gain in the first year of life is currently unknown. The classic predictors of obesity in later childhood and adulthood, such as physical activity, sedentary behavior, and dietary composition, do not translate directly to infancy. In addition, the accumulating body of evidence does not provide robust support for protective effects of breastfeeding or delayed introduction of solid foods. Interventionists have therefore most recently focused efforts on the notion that mothers misread infant hunger and satiety cues and as a result, over-feed them. Underlying these interventions is the implied hypothesis that infants would naturally accurately self-regulate caloric intake if mothers could accurately read the infant's cues. In this application we propose an alternate conceptual model, however. Specifically, we propose that some individuals have greater hunger than others, and that these inter-individual differences are detectable and have high stability from early infancy. We propose that in order to develop effective interventions to prevent rapid weight gain in infancy, there is a need to first understand infant hunger. The goal of this application is to bettr define distinct domains of infant eating behavior, their measurement, their development over time, their correlates, and their association with rapid infant weight gain. Ultimately, the work wll contribute to the basic scientific understanding of the ontogeny of appetitive drive very early in the life span and lead to the development of more precise and targeted interventions to shape infant appetitive drive. We will recruit a cohort of 325 infants in the newborn period and measure, using unique observational methods and maternal report measures, eating behavior constructs including sucking vigor, satiety responsiveness, hedonic response to sweet taste, soothing effect of sucrose, persistent working for food, and eating in the absence of hunger. Our specific aims therefore are: Aim 1: To identify domains of infant eating behavior and their trajectories of development over the first year of life. Aim 2: To test the hypothesis that specifi trajectories of specific infant eating behaviors are associated with more rapid rates of infant weight gain. Aim 3: To test the hypothesis that the infant eating behavior domains linked with rapid rates of weight gain are predicted by infant temperament; psychosocial stress; and maternal obesity, eating behavior, and diet.